Build an MVP for AI Personal Tutor Web & Mobile App for NEET PG Students
Budget: $125 – $250 USD
Looking to develop a secure AI-powered personal tutor web app tailored for NEET PG students in "Tutor-Led Study Mode." The platform will allow students to log in and receive proactive, personalized study guidance based on proprietary notes and their learning history.
Key Features:
- Core Flow: AI analyzes student progress (completed topics, incorrect answers, review gaps) to suggest high-yield topics from secure proprietary notes stored in a vector database (e.g., Supabase or Pinecone).
- AI Functionality: AI retrieves relevant note chunks, teaches briefly, tests with 1–2 MCQs, provides feedback, updates progress logs, and repeats the cycle. If content is unavailable, AI informs the student accordingly.
- Dashboard: Displays "Today’s Plan," chat-style interaction with the AI tutor, and a progress tracker showing mastered topics, weak areas, and review schedules.
- Mode Toggle: Switch between Tutor-Led Study Mode (default) or Ask Anything mode.
- Admin Panel: Upload notes, preprocess them into chunks with metadata (subject/topic/year), trigger embedding into vector DB, and view analytics.
- Mobile-Friendly UI & Secure API Key Storage
- Pages Required: Login/Signup, Dashboard, Study Session Interface, Progress Tracker, Profile/Settings
APIs & Data Management:
- Use Bubble API Connector for OpenAI GPT-4o integration and vector DB search.
- Data includes user profiles, Q&A history, mastery scores, and review schedules.
Additional Notes:
- Limit queries to 50/day; block bulk content requests.
- Ensure proprietary notes are secure and non-downloadable.
Key Features:
- Core Flow: AI analyzes student progress (completed topics, incorrect answers, review gaps) to suggest high-yield topics from secure proprietary notes stored in a vector database (e.g., Supabase or Pinecone).
- AI Functionality: AI retrieves relevant note chunks, teaches briefly, tests with 1–2 MCQs, provides feedback, updates progress logs, and repeats the cycle. If content is unavailable, AI informs the student accordingly.
- Dashboard: Displays "Today’s Plan," chat-style interaction with the AI tutor, and a progress tracker showing mastered topics, weak areas, and review schedules.
- Mode Toggle: Switch between Tutor-Led Study Mode (default) or Ask Anything mode.
- Admin Panel: Upload notes, preprocess them into chunks with metadata (subject/topic/year), trigger embedding into vector DB, and view analytics.
- Mobile-Friendly UI & Secure API Key Storage
- Pages Required: Login/Signup, Dashboard, Study Session Interface, Progress Tracker, Profile/Settings
APIs & Data Management:
- Use Bubble API Connector for OpenAI GPT-4o integration and vector DB search.
- Data includes user profiles, Q&A history, mastery scores, and review schedules.
Additional Notes:
- Limit queries to 50/day; block bulk content requests.
- Ensure proprietary notes are secure and non-downloadable.